38 



SCIENCE. 



[Vol. IX., No. 206 



no significance. The other four pots contained 

 the same sand manured with bone-dust, and these 

 showed an unmistakable loss of nitrogen. This loss, 

 of course, was from the manure rather than from 

 the soil, and it seems probable that it was due to 

 the loss of nitrogen in the free state during decay 

 which has been shown to occur by Reiset, Lawes 

 and Gilbert, Konig and Kieson, Dietzell, Morgen, 

 and others, including the writer. At the same 

 time, these results show that this loss may take 

 place under the circumstances in which organic 

 matter exists in the soil or in the added manure. 

 Strecker observed that the loss was less when the 

 soil was stirred on the surface than when undis- 

 turbed, and greater in the sun than in the shade. 

 He explains the former fact by the hypothesis 

 that the loosened soil absorbed ammonia from the 

 air more freely than the compact one, and thus 

 made good part of the loss just noted. 



The experiments of 1884 were made partly in 

 glass pots, and partly in zinc boxes. Both stood 

 under cover, protected from both rain and dew. 

 Some were filled with sand, and some with gar- 

 den-soil. As before, lupines and oats were used 

 as experimental plants, and pots were also left 

 without plants for the purpose of observing the 

 loss of nitrogen noted in the previous year's ex- 

 periments. 



Strecker's principal conclusions from his results 

 were as follows : — 



1. A naked soil exhales during the summer 

 considerable quantities of nitrogen. The loss is 

 greater from compact than from stirred soil. 

 The results of the experiments of 1884 upon this 

 point were of the same character as those of 1883 ; 

 that is, the results in the sand alone are of no 

 significance, while those in the manured sand 

 show in reality a loss of nitrogen by the manure. 

 In addition to this, however, one of the pots with 

 garden-soil showed an unmistakable loss of ni- 

 trogen. 



2. If the soil is occupied by oats or lupines, this 

 loss of nitrogen is diminished. Some loss was 

 stiU observed in most cases ; but when lupines 

 were grown in unmanured sand, the results, cal- 

 culated on the basis of the minimum percentage 

 of nitrogen originally found in the sand, showed a 

 gain of nitrogen by the soil and roots. An un- 

 mistakable increase of the nitrogen of soil and 

 plant over that of soil and seed was noted in 

 several of these trials in unmanured sand. 



3. In all cases in which the soil was tolerably 

 rich in nitrogen, less nitrogen was found in it 

 after the growth of a crop and the removal of the 

 aerial portions than was present at the beginning 

 of the experiment : in other words, there was no 

 enrichment of the soil. 



4. No essential difference was observed between 

 lupines and oats. Both drew their supply of ni- 

 trogen from the soil, and, in most if not all cases, 

 left it poorer than they found it. 



It will be seen that Strecker's experiments give 

 little countenance to any hypothesis of a gain of 

 nitrogen from the atmosphere. In this respect 

 they differ from the results reported by Atwater.* 



The latter experimented upon peas grown in 

 sand and watered with a solution of plant-food, 

 and found in nearly every case much more ni- 

 trogen in soil and plant than was supplied in seed 

 and nutritive solution. His results, however, do 

 not bear directly upon the question under discus- 

 sion, because he removed the whole plant, in- 

 cluding the roots, from the soil, and determined 

 only the total nitrogen in roots and tops and the 

 residual nitrogen of the soil. It would seem, 

 however, that, if plants can gain so large a pro- 

 portion (up to 50 per cent) of their nitrogen from 

 the air as they did in these experiments, they 

 might very well enrich the soil in nitrogen through 

 their roots and stubble. Strecker's experiments 

 are very interesting as regards the relations of 

 soil and plant to the nitrogen supplies of the 

 atmosphere, but they are entirely inadequate to 

 explain the functions of ' enriching crops ' in agri- 

 culture. Pot experiments, while they permit any 

 exchange of nitrogen between crop and atmos- 

 phere to be accurately observed, practically as- 

 sume that the soil ends at the depth of ten or 

 twelve inches, and take no account of the subsoil 

 as a source of nitrogen. They thus ignore a 

 factor of great importance, and one which aflfects 

 the question in two distinct ways. In the first 

 place, large amounts of nitrates may escape into 

 the subsoil with the drainage-water. I have dis- 

 cussed in an earlier article {Science, iii. No. 48), 

 the results of experiments by Lawes and Gilbert 

 and by Deherain, bearing on this subject, and 

 have shown that the deep-rooting leguminosae, 

 which have a long growing-season, have an im- 

 portant function in arresting these nitrates, and 

 storing them up in an insoluble form, to be set 

 free again gradually for the use of a succeeding 

 crop. According to Lawes and Gilbert, it is at 

 least probable that the roots of clover in some 

 way serve to convey the nitric ferment into the 

 subsoil (which is naturally nearly destitute of it), 

 and thus indirectly convert the insoluble nitrogen 

 compounds there present into nitrates, which they 

 then proceed to assimilate. 



In the second place, it would appear that clover 



and similar deep-rooting plants may bring up 



nitrogen from the subsoil and deposit it in their 



upper roots and stubble. While the soil as a 



1 Amer. chem.journ, vl. 365. 



